Calculates qvalues on the subset PSMs.

Calculates qvalues on the subset PSMs.

calculate_fdr(df, score_higher = TRUE)

Arguments

df	dataframe with at least 3 columns: score score assigned to the peptide to spectrum match (PSM). subset TRUE if PSM belongs to the subset in interest, FALSE or otherwise. decoy TRUE if decoy PSM, FALSE otherwise. Additional columns are allowed but ignored. Target and decoy PSMs are assumbed to be from a competitive target decoy database search.
score_higher	TRUE if a higher score means a better PSM.

Value

A data frame containing all columns in ``df''. Following columns are added:

pi_0_cons

conservative estimation of \(\pi_0\).

FDR

estimated subset PSM qvalues calculated according the competitive target decoy approach.

FDR_BH

estimated subset PSM qvalues calculated according the Benjamini Hochbergh procedure. When provided, non-subset decoy PSMs are used to stabilize estimates in small subsets

FDR_stable

estimated subset PSM qvalues calculated with ``pi_0_cons''. When provided, non-subset decoy PSMs are used to stabilize estimates in small subsets

Examples

## Simulate a dataset with 140 correct target subset PSMs, 60 incorrect target subset PSMS,
## 60 decoy subset PSMs and 2000 additional decoy PSMs.
set.seed(10)
d = sample_dataset(H1_n = 140,H0_n = 60, decoy_n = 60 ,decoy_large_n = 2000)

## calculate the qvalues in the subset target PSMS according the classical target-decoy approach
## and our more stable estimation method.
calculate_fdr(d)
#> # A tibble: 2,260 x 8
#>       score decoy    H0 subset pi_0_cons   FDR FDR_BH FDR_stable
#>       <dbl> <lgl> <lgl>  <lgl>     <dbl> <dbl>  <dbl>      <dbl>
#> 1  2.752437  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> 2  2.726047  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> 3  2.571727  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> 4  2.672108  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> 5  2.788291  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> 6  2.800673  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> 7  2.592950  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> 8  2.702722  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> 9  2.538533  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> 10 2.716658  TRUE FALSE  FALSE     0.305    NA     NA         NA
#> # ... with 2,250 more rows